64 BULLETIN OF THE BUREAU OF FISHERIES 



as the salinity and temperatures that its native water is a large if not the major 

 constituent of the inflowing current, for it is not abundant even along the continental 

 slope (p. 333, and Huntsman, 1919). 



The cold-water siphonophore Diphyes arctica, which occasionally penetrates 

 the Gulf of Maine (p. 379), does so at about the same level as Eukrohnia (about 50 to 

 150 meters), and it is probable that, like the latter, it journeys with the mixed water, 

 in which we have found it over the slope off Shelburne both in March and in June 

 and off the slope of Georges Bank in July, but not along the Nova Scotian coast. 

 The Eastern Channel is, no doubt, the route by which it enters the gulf, judging from 

 the concentration of the localities of capture along the eastern slope of the gulf basin 

 in March and April, 1920. The ultimate origin of D. arctica is not clear as concerns 

 the Gulf of Maine, for while it was formerly supposed to have been one of the most 

 charactersitic of Arctic indicators, captures of it by the Gauss in deep hauls off Cape 

 Verde (Moser, 1915) suggest that it may also range widely in the cold mid-layers of 

 more southern seas, just as Eukrohnia does, and thus reach the gulf from the inter- 

 mediate depths abreast its mouth. 



Sagitta maxima, the largest of local chastognaths, is perhaps the most useful 

 animal indicator of the deepest stratum of the water entering the gulf via the Eastern 

 Channel, both because its habitat is well known offshore, and because it neither breeds 

 in the gulf nor can long survive there, being unfitted for life in water of low salinity no 

 matter what the temperature (Huntsman, 1919, p. 433). S. maxima is so closely con- 

 fined to depths of 150 meters or deeper, both in the Gulf of Maine and in neighboring 

 parts of the Atlantic Ocean, that its presence anywhere in the inner parts of the gulf 

 is unmistakable evidence of the existence of an inflowing current then, or shortly 

 previous, and close to the bottom of the trough. The locality records for S. maxima 

 are concentrated correspondingly in the Eastern Channel, in its immediate debouehe- 

 ment into the general basin of the gulf, and thence northward along its eastern trough 

 as far as the Grand Manan deep, on the one hand, and in the deepest part of the 

 western basin, on the other. As might be expected from its faunistic status, S. 

 maxima is no more periodic (seasonally) than Eukrohnia in its occurrence in the gulf; 

 but although specimens drift in more or less constantly throughout the year, it has 

 invariably been so sparsely represented in hauls made within the gulf, contrasted 

 with considerable abundance at 200 to 300 meters along the continental slope to the 

 east and north, that the indraft can tap only the uppermost levels of its natural 

 habitat offshore at any season. 



The lines of dispersal followed, respectively, by Sagitta serratodentata, Eukrohnia, 

 and S. maxima within the gulf correspond closely with the dominant drift of water at 

 as many levels — that is, surface, mid, and deepest — as made evident by the physical 

 data afforded by temperature and salinity and by drift bottles. Thus, while S. 

 serratodentata not only spreads widely over the offshore parts of the gulf in its season, 

 it also sweeps right around the coast to Massachusetts Bay (which apparently serves 

 more or less as a cul-de-sac for it, as it has for certain drift bottles released in the 

 Bay of Fundy), and Eukrohnia has much the same distribution except that it lives 



